bigbluebutton-Github/bbb-export-annotations/workers/collector.js

172 lines
5.6 KiB
JavaScript

const Logger = require('../lib/utils/logger');
const axios = require('axios').default;
const config = require('../config');
const cp = require('child_process');
const fs = require('fs');
const path = require('path');
const redis = require('redis');
const sanitize = require('sanitize-filename');
const stream = require('stream');
const WorkerStarter = require('../lib/utils/worker-starter');
const {PresAnnStatusMsg} = require('../lib/utils/message-builder');
const {workerData} = require('worker_threads');
const {promisify} = require('util');
const jobId = workerData.jobId;
const logger = new Logger('presAnn Collector');
logger.info(`Collecting job ${jobId}`);
const dropbox = path.join(config.shared.presAnnDropboxDir, jobId);
// Takes the Job from the dropbox
const job = fs.readFileSync(path.join(dropbox, 'job'));
const exportJob = JSON.parse(job);
const jobType = exportJob.jobType;
async function collectAnnotationsFromRedis() {
const client = redis.createClient({
password: config.redis.password,
socket: {
host: config.redis.host,
port: config.redis.port
}
});
client.on('error', (err) => logger.info('Redis Client Error', err));
await client.connect();
const presAnn = await client.hGetAll(jobId);
// Remove annotations from Redis
await client.del(jobId);
const annotations = JSON.stringify(presAnn);
const whiteboard = JSON.parse(annotations);
const pages = JSON.parse(whiteboard.pages);
fs.writeFile(path.join(dropbox, 'whiteboard'), annotations, function(err) {
if (err) {
return logger.error(err);
}
});
// Collect the presentation page files (PDF / PNG / JPEG)
// from the presentation directory
const presFile = path.join(exportJob.presLocation, exportJob.presId);
const pdfFile = `${presFile}.pdf`;
// Message to display conversion progress toast
const statusUpdate = new PresAnnStatusMsg(exportJob);
if (fs.existsSync(pdfFile)) {
for (const p of pages) {
const pageNumber = p.page;
const outputFile = path.join(dropbox, `slide${pageNumber}`);
// CairoSVG doesn't handle transparent SVG and PNG embeds properly,
// e.g., in rasterized text. So textboxes may get a black background
// when downloading/exporting repeatedly. To avoid that, we take slides
// from the uploaded file, but later probe the dimensions from the SVG
// so it matches what was shown in the browser.
const extract_png_from_pdf = [
'-png',
'-f', pageNumber,
'-l', pageNumber,
'-scale-to', config.collector.pngWidthRasterizedSlides,
'-singlefile',
'-cropbox',
pdfFile, outputFile,
];
try {
cp.spawnSync(config.shared.pdftocairo, extract_png_from_pdf, {shell: false});
} catch (error) {
logger.error(`PDFtoCairo failed extracting slide ${pageNumber} in job ${jobId}: ${error.message}`);
statusUpdate.setError();
}
await client.publish(config.redis.channels.publish, statusUpdate.build(pageNumber));
}
} else {
const imageName = 'slide1';
if (fs.existsSync(`${presFile}.png`)) {
fs.copyFileSync(`${presFile}.png`, path.join(dropbox, `${imageName}.png`));
} else if (fs.existsSync(`${presFile}.jpeg`)) {
fs.copyFileSync(`${presFile}.jpeg`, path.join(dropbox, `${imageName}.jpeg`));
} else if (fs.existsSync(`${presFile}.jpg`)) {
// JPG file available: copy changing extension to JPEG
fs.copyFileSync(`${presFile}.jpg`, path.join(dropbox, `${imageName}.jpeg`));
} else {
await client.publish(config.redis.channels.publish, statusUpdate.build());
client.disconnect();
return logger.error(`No PDF, PNG, JPG or JPEG file available for job ${jobId}`);
}
await client.publish(config.redis.channels.publish, statusUpdate.build());
}
client.disconnect();
const process = new WorkerStarter({jobId});
process.process();
}
async function sleep(ms) {
return new Promise((resolve) => {
setTimeout(resolve, ms);
});
}
/** Export shared notes via bbb-pads in the desired format
* @param {Integer} retries - Number of retries to get the shared notes
*/
async function collectSharedNotes(retries = 3) {
/** One of the following formats is supported:
etherpad / html / pdf / txt / doc / odf */
const padId = exportJob.presId;
const notesFormat = 'pdf';
const filename = `${sanitize(exportJob.filename.replace(/\s/g, '_'))}.${notesFormat}`;
const notes_endpoint = `${config.bbbPadsAPI}/p/${padId}/export/${notesFormat}`;
const filePath = path.join(dropbox, filename);
const finishedDownload = promisify(stream.finished);
const writer = fs.createWriteStream(filePath);
try {
const response = await axios({
method: 'GET',
url: notes_endpoint,
responseType: 'stream',
});
response.data.pipe(writer);
await finishedDownload(writer);
} catch (err) {
if (retries > 0 && err?.response?.status == 429) {
// Wait for the bbb-pads API to be available due to rate limiting
const backoff = err.response.headers['retry-after'] * 1000;
logger.info(`Retrying ${jobId} in ${backoff}ms...`);
await sleep(backoff);
return collectSharedNotes(retries - 1);
} else {
logger.error(`Could not download notes in job ${jobId}`);
return;
}
}
const notifier = new WorkerStarter({jobType, jobId, filename});
notifier.notify();
}
switch (jobType) {
case 'PresentationWithAnnotationExportJob': return collectAnnotationsFromRedis();
case 'PresentationWithAnnotationDownloadJob': return collectAnnotationsFromRedis();
case 'PadCaptureJob': return collectSharedNotes();
default: return logger.error(`Unknown job type ${jobType}`);
}